基于高速纹影/阴影成像的流场测速技术研究进展

本文对近年基于纹影/阴影成像的二维和三维速度场测量方法进行了综述，主要内容包括纹影成像的基本原理、硬件设备和测速算法的研究进展。在二维测速方面，介绍了纹影/阴影PIV算法、光流算法及改进算法的原理、适用场景以及优缺点。纹影特性改进光流测速算法可以实现高精度、高空间分辨率的速度场计算，适用范围相对较广。在三维粒子追踪测速方面，主要介绍了层析阴影成像、双视角平行光段阴影成像、双视角汇聚光段阴影成像三种系统的光路设置，并对各自采用的粒子重构和追踪算法进行了比较。双视角阴影成像系统的光路布置更为简洁，降低了对硬件设备的要求，在高速测量中更具优势。梳理了近年来三维粒子追踪测速算法的发展脉络，重点介绍了“先追踪–后重构”和“时间–空间耦合”的双视角三维粒子追踪测速算法。时间–空间耦合的三维粒子追踪测速算法充分利用了时间和空间信息，将时序信息引入立体匹配过程中，显著提升了双视角阴影成像系统在粒子图像密度较高时的重构正确率和追踪准确率，其整体性能优于多种人工智能算法。测速算法在上述方面取得的研究进展，结合短曝光、高帧频的图像采集优势，使得纹影/阴影成像成为一种新型的高帧频、高精度的速度测量技术，在复杂湍流及高瞬态流场实验研究中具有广泛的应用前景。

A review on flow field velocimetry based on high-speed schlieren/shadowgraph systems

The 2-Dimensional （2D） and 3-Dimensional （3D） velocimetry based on schlieren/shadowgraph methods are reviewed in this article. The main content includes the basic optical setups and principles of schlieren and shadowgraph systems, as well as the velocimetry algorithms. For 2D measurement, there are mainly two types of velocimetry algorithms: one is cross-correlation algorithm adopted by PIV, while the other is the optical flow method. The basic formulas, advantages and limitations are introduced comparatively. A recent developed schlieren motion algorithm can provide high accuracy and dense estimation, which is promising and applicable in a wide range of applications. The 3D reconstruction and particle tracking algorithms highly rely on the systems. In this review, three different setups are introduced, including tomographic shadowgraphy, two-view collimated light path shadowgraphy and two-view converging path shadowgraphy. The two-view systems are more concise in setup, requiring less equipment, which are advantageous for high-speed measurements. The 3D particle tracking algorithms of two-view systems are introduced, while the main focus is placed on the image space-based tracking algorithms and the spatial-temporal tracking methods. The latter introduces the temporal predictions into the stereo matching process. The particle reconstruction and tracking correctness in dense particle situations is improved significantly by using the strongly coupled spatial and temporal constraints for optimisation. Its performance is superior to several artificial intelligence methods. The progress of the velocimetry algorithms, together with the imaging advantages of short exposure and high-frequency framing rate, has promoted schlieren/shadowgraph from conventional flow visualization to advanced velocimetry techniques, which can play a role for experimental study in a wide range of complex turbulent and transient flow conditions.